Assignment 2
Due Feb 23, 2021
Goal/Learning Outcomes for students
Understanding the building blocks of a deep learning pipeline- data loading, model files, training/inference, evaluation. Most online tutorials teach how to run MNIST with a standard model architecture (say, ResNet). However, what if you want to modify the model to add a new feature? Run on a new dataset? Measure using a different evaluation metric? This assignment takes a first step towards setting you up to approach these questions. Over the course of the semester, you can build on this code-base and use it to work on your own project.
Deliverables
Part 1: All students need to declare their project as part of Assignment 2! Please submit this on canvas.
Part 2: If you’re in Track A: Run Assignment2.ipynb to the end. Download it as an HTML and submit it to canvas. Updating your forked repository is not a requirement for getting full credit.
If you’re in Track B: You don’t need to submit anything in this assignment, if you submitted evidence of ML experience in Assignment 1.